ABSTRACT Psychotic disorders are severe, debilitating illnesses with limited treatment options. Identification of individuals who are at elevated risk for developing psychotic disorders is a critical first step for prevention. Prior research has established that those individuals at clinical high risk (CHR) who go on to develop a full-blown psychotic disorder (converters), as a group, tend to exhibit excess dopamine in the nigro-striatal pathway, as measured by PET. But this finding is unlikely to help identify individual subjects given its low predictive value and the practical limitations associated with PET (e.g., radiation exposure). Previous research has also established that a combination of widely available clinical variables ?the NAPLS2 calculator? can predict risk of conversion with moderate predictive value. Here, we propose to use neuromelanin-sensitive MRI (NM-MRI), a novel, easy-to- acquire, non-invasive MRI technique that can be safely used in pediatric individuals and that provides a proxy for psychosis-related nigro-striatal dopamine excess, as a predictive biomarker for risk of conversion in CHR individuals. Furthermore, we aim to combine this objective biomarker with the clinical (subjective) information in the NAPLS2 calculator to test whether this biomarker can improve the accuracy of individual risk prediction beyond that achieved by clinical information alone, a critical test of the potential clinical utility of a biomarker that previous imaging studies in CHR populations have largely ignored. Thus, we aim to combine the strengths of an easy-to-acquire, objective MRI biomarker tapping into the pathophysiology of psychosis and those of an established risk algorithm based on clinical data to more accurately identify individuals at risk for developing full-blown psychotic disorders and predict time to conversion. Specifically, and as supported by our preliminary data, we first aim to determine whether baseline NM-MRI of the substantia nigra, pars compacta (SNc) can reveal abnormally increased signal specifically in those CHR individuals with more severe attenuated psychotic symptoms. Second, we aim to determine whether baseline NM-MRI SNc signal is particularly elevated in CHR converters compared to non-converters and to sociodemographically matched healthy controls. Third, we aim to assess whether NM-MRI SNc signal can improve the accuracy of risk predictions over and above those derived from the NAPLS2 calculator. If successful, this proposal will thus establish the potential clinical utility of a novel MRI biomarker that can be adopted widely and used safely in pediatric and non-pediatric populations to enhance risk predictions for the development of psychosis and potentially to monitor treatment and aid in personalized treatment selection.